Delay Measurement of Concurrent Video Streams Analysis in Autonomous Vehicle DCUs

Abstract

Autonomous vehicles (AVs) are one of the most innovative research fields. The advanced AVs leverage a multitude of sensors, including LiDAR, radar, cameras, and ultrasonic sensors, in combination with Hybrid Vehicle-to-Everything (V2X) technology, to achieve a comprehensive understanding of their environment, enabling precise navigation, real-time communication with infrastructure and other vehicles, and enhancing overall safety and efficiency on the roads.

The Networked Intelligence Laboratory is working on the Hybrid V2X project. The project focuses on the platform that includes multiple interconnected high-computing hardware units with advanced cameras. Multiple data concentrator units (DCUs) were built for the project to perform object-detection analysis of several high-definition video streams. One of the main objectives of the Hybrid V2X project is to maintain the optimal performance of DCUs with minimal system failures.

Maintaining low latency of the DCU's performance is crucial since object detection is performed with real-time data. For that purpose, the remote monitoring tool is built that analyzes the delay in performance of multiple DCUs simultaneously. This work presents the delay measurement tool for multiple locally distributed DCUs used in the Hybrid V2X project.